Tuberculosis infects much of the world's population and is responsible for millions of deaths annually. The causative organism, Mycobacterium tuberculosis, has been recognized for over a century, but little is known about the molecular mechanisms used by this bacterium to cause disease. I propose to use three methods we have recently developed in conjunction with the M tuberculosis genomic sequence to determine which genes are required by M tuberculosis to survive both in vitro and in vivo. First, we have developed a new transposon to perform saturating mutagenesis in M tuberculosis. Second, we have made a DNA microarray with which we can measure hybridization to each M tuberculosis open reading frame. Third, we have developed transposon junction hybridization (TJH), a method for mapping the sites of transposon insertions in large pools of mutants using a DNA microarray. We propose to use TJH to compare the genes required for M tuberculosis in vitro growth with those needed to survive in an animal. We will use a variation of TJH, differential length hybridization, to identify the complete set of genes that are essential for growth in defined media. We will also sequence several thousand clones from an insertion mutant library to produce a bank of defined mutants. This will allow us to test individual strains that contain mutations in candidate virulence genes identified by TJH. Since pathogens coordinately regulate expression of virulence genes, we will focus on regulatory genes required for infection and determine which downstream genes they control. This will enable us to identify both genes required for survival and for causing disease. Identification of genes important in infection should lead to the development of new strategies of tuberculosis treatment and prevention.